This comment discusses information hazards, but not in much detail.
“Don’t link to possible information hazards on Less Wrong without clear warning signs.” — Eliezer, in the previous open thread.
“Information hazard” is a Nick Bostrom coinage. The previous discussion of this seems to have focused on what Bostrom calls “psychological reaction hazard” — information that will make (at least some) people unhappy by thinking about it. Going through Bostrom’s paper on the subject, I wonder if these other sorts of information hazards should also be avoided here:
Distraction hazards — addictive products, games, etc.; especially those that have been optimized to be so. Examples: Links to video games; musical earworms; discussions of addictive drug use; porn.
Role model hazards — discussions of people doing harmful things; bad examples that readers might imitate. Examples: Talking about suicide and thoughts leading to it; fatalistic discussion of bad habits.
Biasing hazards — information that amplifies existing biased beliefs. Examples skipped to avoid a distracting political discussion here.
Embarrassment hazard — discussions of embarrassing things happening to people in the community. Examples: Links to scandalous or distorted stories about members of the community; gossip in general.
Another thing that seems to fit this pattern that I have seen elsewhere is a Trigger Warning, which is used before people discuss something like rape, discrimination, etc… which can remind people who have experienced those about it, causing some additional trauma from the event.
Has anyone here ever decided not to read something because it had a trigger warning? I can’t imagine doing so myself, but that may be the typical mind fallacy.
Has anyone here ever decided not to read something because it had a trigger warning? I can’t imagine doing so myself, but that may be the typical mind fallacy.
I have chosen not to consume media (including but not limited to text) because of an explicit trigger warning. Not often, though; most trigger warnings relate to topics I don’t have trauma about.
More often, I have chosen to defer consuming media because of an explicit trigger warning, to a time and place when/where emotional reactions are more appropriate.
I have consumed media in the absence of such warnings that, had such a warning been present, I would have likely chosen to defer. In some cases this has had consequences I would have preferred to avoid.
I haven’t, but I think that were trigger warnings are appropriate is in things that hurt a few people disproportionately. If something hurts everyone that reads it you shouldn’t write it at all, and if it hurts no one more than it is worth it isn’t a case for trigger warnings. But if it is something that needs to be said to many people, and there is a significant group (perhaps those that have had a certain experience) who would suffer a lot from reading it, then you put a trigger warning that would be recognized by that group at the top.
TLDR If most people never care about trigger warnings, then they might work as intended.
Trigger warnings are stupid in general, I think they do more harm than good.
Even people who fear being negatively affected will mostly read the content, if only because forbidden fruit are the sweetest and because they are curious. The trigger warning will then already have put them in a frame of mind in which they expect a bad emotional impact of some sort—clearly predisposing them to react much worse than if there had been no trigger warning in the first place.
I concede that some people may in fact heed trigger warnings and not read the content, but an overall utility calculation would probably favor no trigger warnings at all.
Even people who fear being negatively affected will mostly read the content, if only because forbidden fruit are the sweetest and because they are curious.
Probably, people for whom that is true (while constituting probably the majority of regular Internet users) are not the same people as those for whom trigger warnings are written. See e.g. this discussion about the relationship between the openness personality trait and the memetic analogue of parasite load.
I have chosen not to Google something that I was warned would involve seeing particularly horrific images. I imagine that if said topic was put in blog post form with a trigger warning up the top, I would probably choose not to read it.
EDIT: It’s probably worth adding that I adopted this policy after discovering the hard way that there are things out there I would really prefer not to see/hear about.
I haven’t, but I have never experienced a serious trauma that I don’t want to be reminded to me, so I’m not the kind of person that people who write trigger warnings are thinking about.
Agreed — Bostrom’s classification “psychological reaction hazard” seems like it should include “trigger” as a subset — both the original sense of “PTSD trigger” and the more general sense that seems popular today, which might be expanded as “information that will remind you of something that it hurts to be reminded of.”
As for distraction hazards, I have often seen links to TvTropes been posted with a warning sign, both here and in other sites. (Sometimes a plain “Warning: TvTropes link”, sometimes a more teasing “Warning: do not click link unless you have hours to spare today”.)
Or “Warning: Daily Mail” (or other sites working on the click-troll business model): linking to a site your readers may object to feeding even with a click. Knowledge hazard, in that even when such sites are more right than their usual level they tend to be wrong.
Why stop there? Employment hazard (NSFW), Copyright hazard (link to torrent, sharing site or a paper copied from behind a paywall), Relationship hazard (picture of a gorgeous guy/girl), dieting hazard (discussion of what goes well with bacon)...
Well, the ones I mentioned are drawn from Bostrom’s paper (although they aren’t all of his categories). Eliezer seemed to be specifically discouraging a class of psychological reaction hazards while using the more general term “information hazard” to do it; I thought to inquire into what folks thought of other classes of information hazard.
This comment discusses information hazards, but not in much detail.
“Don’t link to possible information hazards on Less Wrong without clear warning signs.”
— Eliezer, in the previous open thread.
“Information hazard” is a Nick Bostrom coinage. The previous discussion of this seems to have focused on what Bostrom calls “psychological reaction hazard” — information that will make (at least some) people unhappy by thinking about it. Going through Bostrom’s paper on the subject, I wonder if these other sorts of information hazards should also be avoided here:
Distraction hazards — addictive products, games, etc.; especially those that have been optimized to be so. Examples: Links to video games; musical earworms; discussions of addictive drug use; porn.
Role model hazards — discussions of people doing harmful things; bad examples that readers might imitate. Examples: Talking about suicide and thoughts leading to it; fatalistic discussion of bad habits.
Biasing hazards — information that amplifies existing biased beliefs. Examples skipped to avoid a distracting political discussion here.
Embarrassment hazard — discussions of embarrassing things happening to people in the community. Examples: Links to scandalous or distorted stories about members of the community; gossip in general.
Another thing that seems to fit this pattern that I have seen elsewhere is a Trigger Warning, which is used before people discuss something like rape, discrimination, etc… which can remind people who have experienced those about it, causing some additional trauma from the event.
Has anyone here ever decided not to read something because it had a trigger warning? I can’t imagine doing so myself, but that may be the typical mind fallacy.
EDIT: People do use the warnings. Good to know.
I have chosen not to consume media (including but not limited to text) because of an explicit trigger warning. Not often, though; most trigger warnings relate to topics I don’t have trauma about.
More often, I have chosen to defer consuming media because of an explicit trigger warning, to a time and place when/where emotional reactions are more appropriate.
I have consumed media in the absence of such warnings that, had such a warning been present, I would have likely chosen to defer. In some cases this has had consequences I would have preferred to avoid.
I haven’t, but I think that were trigger warnings are appropriate is in things that hurt a few people disproportionately. If something hurts everyone that reads it you shouldn’t write it at all, and if it hurts no one more than it is worth it isn’t a case for trigger warnings. But if it is something that needs to be said to many people, and there is a significant group (perhaps those that have had a certain experience) who would suffer a lot from reading it, then you put a trigger warning that would be recognized by that group at the top.
TLDR If most people never care about trigger warnings, then they might work as intended.
Trigger warnings are stupid in general, I think they do more harm than good.
Even people who fear being negatively affected will mostly read the content, if only because forbidden fruit are the sweetest and because they are curious. The trigger warning will then already have put them in a frame of mind in which they expect a bad emotional impact of some sort—clearly predisposing them to react much worse than if there had been no trigger warning in the first place.
I concede that some people may in fact heed trigger warnings and not read the content, but an overall utility calculation would probably favor no trigger warnings at all.
Probably, people for whom that is true (while constituting probably the majority of regular Internet users) are not the same people as those for whom trigger warnings are written. See e.g. this discussion about the relationship between the openness personality trait and the memetic analogue of parasite load.
I have chosen not to Google something that I was warned would involve seeing particularly horrific images. I imagine that if said topic was put in blog post form with a trigger warning up the top, I would probably choose not to read it.
EDIT: It’s probably worth adding that I adopted this policy after discovering the hard way that there are things out there I would really prefer not to see/hear about.
I’ve decided not to listen to some radio segments because of such warnings. Similar principle.
Have you had an experience that might cause you to be triggered by the kind of thing that gets trigger warnings?
I haven’t, but I have never experienced a serious trauma that I don’t want to be reminded to me, so I’m not the kind of person that people who write trigger warnings are thinking about.
I know a person who chose not to read something (MAX Punisher #1) based on my warning of explicit sexual violence.
Anecdotal and incomplete, but most of an example case...
Agreed — Bostrom’s classification “psychological reaction hazard” seems like it should include “trigger” as a subset — both the original sense of “PTSD trigger” and the more general sense that seems popular today, which might be expanded as “information that will remind you of something that it hurts to be reminded of.”
As for distraction hazards, I have often seen links to TvTropes been posted with a warning sign, both here and in other sites. (Sometimes a plain “Warning: TvTropes link”, sometimes a more teasing “Warning: do not click link unless you have hours to spare today”.)
Or “Warning: Daily Mail” (or other sites working on the click-troll business model): linking to a site your readers may object to feeding even with a click. Knowledge hazard, in that even when such sites are more right than their usual level they tend to be wrong.
I wish links to Cracked.com also had a similar warning. (Well, now that I have LeechBlock installed that’s no longer so much of an issue, but still.)
Why stop there? Employment hazard (NSFW), Copyright hazard (link to torrent, sharing site or a paper copied from behind a paywall), Relationship hazard (picture of a gorgeous guy/girl), dieting hazard (discussion of what goes well with bacon)...
Well, the ones I mentioned are drawn from Bostrom’s paper (although they aren’t all of his categories). Eliezer seemed to be specifically discouraging a class of psychological reaction hazards while using the more general term “information hazard” to do it; I thought to inquire into what folks thought of other classes of information hazard.